22 PROFESSOR CONNELL ON A 



it with every care, I then obtained a result in excess, and therefore am inclined 

 to think, that the tendency of that instrument is so decidedly in excess, although 

 certainly to only a small extent, that I doubt the perfect accuracy of any operation 

 with it in deficiency. I however retain the average, as observation gave it. With 

 regard to the instrument described in this paper, it affords, in sixteen observations, 

 thirteen results in deficiency with reference to the same standard, and three in 

 excess, the average of the whole being 0°95 in deficiency. 



These observations, I think, entitle me to ask that any one wishing to make 

 an experimental estimate of the value of the indications of this instrument, shall 

 not assume Daniell's instrument as giving indications which can be rigidly con- 

 sidered as standards of comparison. It is by Dalton's method that I should 

 wish it to be tried ; and, I may mention, that in using this method I employed two 

 German becker glasses, 3^ inches deep by 2\ wide, and added a mixture of equal 

 parts of nitre and sal-ammoniac, in powder, to the water, till it shewed dew ; and 

 then, by transference of the liquid from one to the other, carefully cleansing and 

 drying the surfaces, noted the point of disappearance of moisture, and took a mean 

 between that and the last observation shewing dew, provided the difference did not 

 exceed 1° or 1^° ; being satisfied that there were many more chances that the 

 true dew point lay in this interval, than at the first temperature at which no dew 

 was noticed. The difference, however, resulting from this mode of observation 

 in the noted result scarcely in any case exceeds o, 5 F., and is usually much less* 



I think it preferable, on the whole, to leave the little bottle with its proper 

 brass surface duly polished, rather than to have it gilt ; at least I think that every 

 one will find this to be the case, who will take care always to have it preserved 

 quite bright and clean, which is easily done by the use, when necessary, of ordi- 

 nary brass polishing paste, f For those who would wish to have it always in a 

 state fit for use, with little or no trouble, although at the cost of a little polish of 

 surface, it will be best to have the surface gilt. Although not quite so delicate in 

 its indications, it will be sufficiently so for use. This is a matter which may be 

 safely left to the option of individuals. Possibly a bottle of polished silver might, 

 on the whole, be better than either, but I do not think it likely that any advan- 

 tage in this way would be worth the additional expense. This method is, how- 

 ever, open to any one who wishes it. 



* I occasionally noticed that after adding the salts, the first deposition of moisture could not 

 he noticed before the surface of the glass had imperceptibly become quite moist, and the temperature 

 much below the true dew point ; but this, of course, was easily corrected, by transferring the liquid to 

 a dry vessel, and proceeding in the usual way. 



•j" The manner recommended to me is to scrape off with a knife from a piece of good and light 

 rotten-stone, some very fine powder, to place this on a piece of woollen cloth, such as the rind of 

 broad cloth, to mix this with a little olive-oil, and rub the bottle with this mixture, then to rub it with 

 a piece of cotton cloth, on which a little of the fine powdered rotten-stone has been laid, and to finish 

 by rubbing with a piece of soft cloth, without either powder or oil, till a bright surface is obtained. 

 A piece of chamois leather is also kept with which to clean it, when the paste is not used. 



